Introduction to JupyterLab
- David Palmquist
- Library Systems Team, CSUF Pollak Library
- Chat Assistance
- Tu-an Nguyen, Analyst/Programmer
- Christina Cagu, Student Assistant
What will be touched on
- What is JupyterLab
- Where can you run it
- Interface basics
- Resources
- Q & A
Poll
Before rolling any further, lets take a moment for a short poll to gauge audience familiarity.
What is JupyterLab?
- Browser-based interactive environment
- Combines functional code, data exploration, and presentation in a single portable file
- Supports Python, R, Julia, and more
- Ideal for data science, research, and teaching
JupyterLab: power of this, without all the clutter
Or if you prefer, much of this, less clutter
Vanilla out of the box, looks something like this
And you can even go inception and Juptyer like this
Where can you run Jupyter
Jupyter on Your Hardware
- Requires more setup effort
- Limited by your hardware, but keeps data local
- Example of launching local
- Notice: URL location
Jupyter Cloud-Based Options
- Multiple platforms available:
- Google Colab
- Kaggle Notebooks
- Azure Notebooks
- Binder
- GitHub Codespaces
- JupyterHub (e.g., Titan Computing Hub)
Today we will focus on univeral tips for the common experience
- In the interest of time, lets all work from this jupyter lite
example: https://jupyter.org/try-jupyter/lab/
- jupyter lite is limited since it runs in a browswer sandbox
- requires no installation
- limited persistence to browser cache / network options
- jupyter lite is limited since it runs in a browswer sandbox
JupyterLab Interface
- Menu Bar: File, Edit, View, Run options
- Left Sidebar: File browser, running kernels, extensions
- Main Work Area: Notebooks, terminals, text editors
Operations, Tips, and Shortcuts 1
- Create a notebook
- Open a notebook
- Create a Code Cell
- Run cells
Operations, Tips, and Shortcuts 2
- Markdown cell
- Headings
- Lists
- ordered
- unordered
- Links
- Images
Operations, Tips, and Shortcuts 3
- Use Command Mode and keyboard shortcuts
- [esc]
- [a],[b],[x],[z],[m],[y]
- magic commands
%%time sum(i**2 for i in range(1, 50000001))
Kernels
- Interrupt
- Restart
- while True: x=5
Where to Go from Here
- Experiment with Jupyter to assess its value to your process
- Explore what your colleagues have already done:
- Berkely Data Science Modules
- [Gallery of Notebooks (api-issue)]
- Explore what your colleagues have already done:
- Learn to install and use Jupyter locally
- Explore Carpentries workshops
- Self-study via LinkedIn Learning
Resources
Q & A
Jupyter is an open source tool. Powerful in isolation and even more so in collaboration. It helps limit cognitive load in both generation and consumption of knowledge. In my opinion, the learning curve is worth it, so add it to your toolbox today. Figure out how many things look like a nail tomorrow :)
What questions should we explore?